Examining Response Styles and Their Impact on Psychological Testing Outcomes: A Mixture IRT (Item Response Theory) Modeling Approach (Poster 2)

• In Event: Sunday Poster Session 9:35 am
  In Poster Session: Applied Statistical Methods

Abstract

This study investigates the impact of response styles on psychological testing outcomes using complicated mixed Item Response Theory (IRT) models. Synthetic data with diverse response patterns, including dominant and ideal-response styles, is generated to reflect real-world variability. The mixture IRT models successfully identify latent response styles, such as Dominant Responders, Ideal-Response Seekers, and Balanced Responders. Significant differences in test scores are observed between these groups, with Dominant Responders showing inflated scores and Ideal-Response Seekers displaying score deflation. The study highlights the importance of considering response styles in psychological testing and their implications for test validity. By disentangling latent response styles from traditional IRT factors, researchers can improve the precision and accuracy of psychological assessments.

Author

• Fatih Ozkan, Baylor University